基于优化变分模态分解和核极限学习机的集装箱吞吐量预测

doi:10.11772/j.issn.1001-9081.2021050816

《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (8): 2333-2342.DOI: 10.11772/j.issn.1001-9081.2021050816

• 人工智能 • 上一篇

基于优化变分模态分解和核极限学习机的集装箱吞吐量预测

张丰婷¹, 杨菊花¹(), 任金荟², 金坤¹

^1.兰州交通大学交通运输学院, 兰州 730070
^2.中国铁路兰州局集团有限公司兰州货运中心安全生产部, 兰州 730030

收稿日期:2021-05-19 修回日期:2021-09-25 接受日期:2021-09-26 发布日期:2021-09-25 出版日期:2022-08-10
通讯作者: 杨菊花
作者简介:张丰婷（1996—），女，甘肃武威人，硕士研究生，主要研究方向：交通运输规划与管理、机器学习；
杨菊花（1978—），女，甘肃天水人，副教授，博士，主要研究方向：交通运输规划与管理；
任金荟（1991—），男，甘肃定西人，主要研究方向：交通运输；
金坤（1996—），男，甘肃白银人，硕士研究生，主要研究方向：机器学习、人工智能。
基金资助:
甘肃省自然科学基金资助项目(21JR7RA287);甘肃省教育厅“双一流”科研重点项目(GSSYLXM-04)

Container throughput prediction based on optimal variational mode decomposition and kernel extreme learning machine

Fengting ZHANG¹, Juhua YANG¹(), Jinhui REN², Kun JIN¹

^1.School of Traffic and Transportation，Lanzhou Jiaotong University，Lanzhou Gansu 730070，China
^2.Safety Production Department of Lanzhou Freight Center，China Railway Lanzhou Bureau Group Company Limited，Lanzhou Gansu 730030，China

Received:2021-05-19 Revised:2021-09-25 Accepted:2021-09-26 Online:2021-09-25 Published:2022-08-10
Contact: Juhua YANG
About author:ZHANG Fengting， born in 1996， M. S. candidate. Her research interests include transportation planning and management， machine learning.
YANG Juhua， born in 1978， Ph. D.， associate professor. Her research interests include transportation planning and management.
REN Jinhui， born in 1991. His research interests include transportation.
JIN Kun， born in 1996， M. S. candidate. His research interests include machine learning， artificial intelligence.
Supported by:
Natural Science Foundation of Gansu Province(21JR7RA287);“Double-First Class” Major Research Program of Educational Department of Gansu Province(GSSYLXM-04)

摘要/Abstract

摘要：

针对港口集装箱吞吐量数据的复杂性特征，提出基于优化变分模态分解（OVMD）和核极限学习机（KELM）的集装箱吞吐量短期混合预测模型。首先，用汉佩尔辨识法（HI）剔除原始时间序列中的异常值，并把预处理之后的序列通过OVMD分解为多个特征明显的子模态。然后，为提高预测效率，将分解后的子模态按照样本熵（SE）值的大小分成高频低幅、中频中幅和低频高幅三类；同时，借助KELM中携带的小波、高斯和线性核函数捕捉具有不同特征子模态的趋势。最后，把所有子模态的预测结果线性相加得到最终的预测结果。以深圳港的月度集装箱吞吐量数据为样本进行实验，所提模型的平均绝对误差（MAE）达到0.914?9，平均绝对百分比误差（MAPE）达到0.199%，均方根误差（RMSE）达到7.886?0，决定系数（R²）为0.994?4。与四种对比模型相比，所提出的模型在预测精度和效率上都具有一定的优势，同时克服了传统互补集成经验模态分解（CEEMD）和集成经验模态分解（EEMD）中容易出现的模态混叠问题以及极限学习机（ELM）中存在过拟合等问题，具有一定的实际应用潜力。

关键词: 集装箱吞吐量预测, 样本熵, 变分模态分解, 核极限学习机, 分解集成预测模型

Abstract:

Aiming at the complexity of port container throughput data， a short-term hybrid prediction model of container throughput based on Optimal Variational Mode Decomposition （OVMD） and Kernel Extreme Learning Machine （KELM） was proposed. Firstly， the outliers were removed by Hampel Identifier （HI） from the original time series， and the preprocessed series was decomposed into several sub-modes with obvious characteristics by OVMD. Then， in order to improve the prediction efficiency， the decomposed sub-modes were divided into three categories according to the values of Sample Entropy （SE）： high frequency low amplitude， medium frequency medium amplitude and low frequency high amplitude. At the same time， the wavelet， Gauss and linear kernel functions carried in KELM were used to capture the trends of sub-modes with different characteristics. Finally， the final prediction result was obtained by linearly adding the prediction results of all sub- modes together. Taking the monthly container throughput data at Shenzhen Port as a sample for empirical research， the proposed model has the Mean Absolute Error （MAE） of 0.914?9， the Mean Absolute Percentage Error （MAPE） of 0.199%， the Root Mean Square Error （RMSE） of 7.886?0 and the coefficient of determination （R²） of 0.994?4. Compared with four comparison models， the proposed model has advantages in prediction accuracy and efficiency. At the same time， it overcomes the mode mixing problem in traditional Complementary Ensemble Empirical Mode Decomposition （CEEMD） and Ensemble Empirical Mode Decomposition （EEMD） as well as overfitting defect in Extreme Learning Machine （ELM）， and has practical application potential.

Key words: container throughput prediction, Sample Entropy (SE), Variational Mode Decomposition (VMD), Kernel Extreme Learning Machine (KELM), decomposition-ensemble prediction model

中图分类号:

U691.71

张丰婷, 杨菊花, 任金荟, 金坤. 基于优化变分模态分解和核极限学习机的集装箱吞吐量预测[J]. 计算机应用, 2022, 42(8): 2333-2342.

Fengting ZHANG, Juhua YANG, Jinhui REN, Kun JIN. Container throughput prediction based on optimal variational mode decomposition and kernel extreme learning machine[J]. Journal of Computer Applications, 2022, 42(8): 2333-2342.

图/表 13

图1 OVMD流程

Fig. 1 Process of OVMD

图2 相点与元素的映射关系

Fig. 2 Mapping relation between phase points and elements

图3 ELM的拓扑结构

Fig. 3 Topological structure of ELM

图4 本文集装箱吞吐量预测模型的框架

Fig. 4 Framework of the proposed container throughput prediction model

图5 深圳港月度集装箱吞吐量

Fig. 5 Monthly container throughput at Shenzhen Port

图6 集装箱吞吐量月度时间序列的ACF分析

Fig. 6 ACF analysis of monthly time series of container throughput

图7 HI离群点处理结果

Fig. 7 HI outlier processing results

图8 OVMD结果

Fig. 8 OVMD results

表1 各IMF的样本熵值

Tab. 1 Sample entropy values of IMFs

分量	样本熵值	分量	样本熵值
IMF1	0.019 5	IMF8	0.282 0
IMF2	0.002 4	IMF9	0.220 5
IMF3	0.148 9	IMF10	0.423 6
IMF4	0.495 3	IMF11	0.539 2
IMF5	0.262 4	IMF12	0.442 9
IMF6	0.317 5	IMF13	0.161 8
IMF7	0.077 2

图9 不同核函数预测对比

Fig. 9 Prediction comparison of different kernel functions

图10 本文模型和其他模型的预测曲线对比

Fig. 10 Prediction curves comparison of the proposed model and other models

表2 不同模型的性能评价指标

Tab. 2 Performance evaluation indicators of different models

模型	MAE↓	MAPE↓/%	RMSE $↓$	R²→1
HI-OVMD-SE-KELM	0.914 9	0.1990	7.8860	0.9944
HI-CEEMD-SE-KELM	6.183 8	1.342 7	51.908 0	0.776 8
HI-EEMD-SE-KELM	5.284 7	1.157 3	44.489 4	0.858 8
HI-OVMD-ARIMA	32.809 1	6.919 4	248.419 5	0.038 0
HI-OVMD-ELM	10.449 9	2.203 2	92.732 7	0.402 6

表2 不同模型的性能评价指标

Tab. 2 Performance evaluation indicators of different models

模型	MAE↓	MAPE↓/%	RMSE $↓$	R²→1
HI-OVMD-SE-KELM	0.914 9	0.1990	7.8860	0.9944
HI-CEEMD-SE-KELM	6.183 8	1.342 7	51.908 0	0.776 8
HI-EEMD-SE-KELM	5.284 7	1.157 3	44.489 4	0.858 8
HI-OVMD-ARIMA	32.809 1	6.919 4	248.419 5	0.038 0
HI-OVMD-ELM	10.449 9	2.203 2	92.732 7	0.402 6

图11 不同模型的预测误差曲线

Fig. 11 Prediction error curves of different models

参考文献 29

1	NOTTEBOOM T. The adaptive capacity of container ports in an era of mega vessels： the case of upstream seaports Antwerp and Hamburg［J］. Journal of Transport Geography， 2016， 54： 295-309. 10.1016/j.jtrangeo.2016.06.002
2	TWRDY E， BATISTA M. Modeling of container throughput in Northern Adriatic ports over the period 1990-2013［J］. Journal of Transport Geography， 2016， 52： 131-142. 10.1016/j.jtrangeo.2016.03.005
3	LI M W， GENG J， HONG W C， et al. A novel approach based on the Gauss-vSVR with a new hybrid evolutionary algorithm and input vector decision method for port throughput forecasting［J］. Neural Computing and Applications， 2017， 28（S1）： 621-640. 10.1007/s00521-016-2396-3
4	VILJOEN N M， JOUBERT J W. The vulnerability of the global container shipping network to targeted link disruption［J］. Physica A： Statistical Mechanics and its Applications， 2016， 462： 396-409. 10.1016/j.physa.2016.06.111
5	CHEN S H， CHEN J N. Forecasting container throughputs at ports using genetic programming［J］. Expert Systems with Applications， 2010， 37（3）： 2054-2058. 10.1016/j.eswa.2009.06.054
6	ZHANG C， HUANG L， ZHAO Z C. Research on combination forecast of port cargo throughput based on time series and causality analysis［J］. Journal of Industrial Engineering and Management， 2013， 6（1）： 124-134. 10.3926/jiem.687
7	杜柏松，艾万政，胡林燕，等.基于优化灰色马尔科夫动态模型的上海港集装箱吞吐量预测［J］.上海海事大学学报， 2021， 42（1）： 76-81.
	DU B S， AI W Z， HU L Y， et al. Container throughput prediction of Shanghai port based on the optimized grey Markov dynamic model［J］. Journal of Shanghai Maritime University， 2021， 42（1）： 76-81.
8	NIU M F， WANG Y F， SUN S L， et al. A novel hybrid decomposition-and-ensemble model based on CEEMD and GWO for short-term PM2.5 concentration forecasting［J］. Atmospheric Environment， 2016， 134： 168-180. 10.1016/j.atmosenv.2016.03.056
9	查茜.港口集装箱吞吐量时间序列预测方法研究［D］.重庆：重庆大学， 2016： 9-17.
	ZHA Q. Research on container throughput time series forecasting methods［D］. Chongqing： Chongqing University， 2016： 9-17.
10	KORDANULI B， BARJAKTAROVIĆ L， JEREMIĆ L， et al. Appraisal of artificial neural network for forecasting of economic parameters［J］. Physica A： Statistical Mechanics and its Applications， 2017， 465： 515-519. 10.1016/j.physa.2016.08.062
11	DIXON M. Sequence classification of the limit order book using recurrent neural networks［J］. Journal of Computational Science， 2018， 24： 277-286. 10.1016/j.jocs.2017.08.018
12	RICHHARIYA B， TANVEER M. EEG signal classification using universum support vector machine［J］. Expert Systems with Applications， 2018， 106： 169-182. 10.1016/j.eswa.2018.03.053
13	ZHANG Z G， KON M A. Wavelet sampling and generalization in neural networks［J］. Neurocomputing， 2017， 267： 36-54. 10.1016/j.neucom.2017.04.054
14	GOSASANG V， CHANDRAPRAKAIKUL W， KIATTISIN S. A comparison of traditional and neural networks forecasting techniques for container throughput at Bangkok Port［J］. The Asian Journal of Shipping and Logistics， 2011， 27（3）： 463-482. 10.1016/s2092-5212(11)80022-2
15	XIE G， WANG S Y， ZHAO Y X， et al. Hybrid approaches based on LSSVR model for container throughput forecasting： a comparative study［J］. Applied Soft Computing， 2013， 13（5）： 2232-2241. 10.1016/j.asoc.2013.02.002
16	GENG J， LI M W， DONG Z H， et al. Port throughput forecasting by MARS-RSVR with chaotic simulated annealing particle swarm optimization algorithm［J］. Neurocomputing， 2015， 147： 239-250. 10.1016/j.neucom.2014.06.070
17	LI B， RONG X W， LI Y B. An improved kernel based extreme learning machine for robot execution failures［J］. The Scientific World Journal， 2014， 2014： No.906546. 10.1155/2014/906546
18	张如九，严磊，何旭辉，等.基于混合方法的风速预测模型研究［J］.铁道科学与工程学报， 2020， 17（7）： 1630-1636.
	ZHANG R J， YAN L， HE X H， et al. Research on wind velocity prediction models based on hybrid methods［J］. Journal of Railway Science and Engineering， 2020， 17（7）： 1630-1636.
19	CHANG Y S， CHIAO H T， ABIMANNAN S， et al. An LSTM-based aggregated model for air pollution forecasting［J］. Atmospheric Pollution Research， 2020， 11（8）： 1451-1463. 10.1016/j.apr.2020.05.015
20	徐玉萍，邓俊翔，蒋泽华.基于组合预测模型的铁路货运量预测研究［J］.铁道科学与工程学报， 2021， 18（1）： 243-249.
	XU Y P， DENG J X， JIANG Z H. Railway freight volume forecasting based on a combined model［J］. Journal of Railway Science and Engineering， 2021， 18（1）： 243-249.
21	DRAGOMIRETSKIY K， ZOSSO D. Variational mode decomposition［J］. IEEE Transactions on Signal Processing， 2014， 62（3）： 531-544. 10.1109/tsp.2013.2288675
22	HUANG G B， ZHU Q Y， SIEW C K. Extreme learning machine： theory and applications［J］. Neurocomputing， 2006， 70（1/2/3）： 489-501. 10.1016/j.neucom.2005.12.126
23	NIU M F， HU Y Y， SUN S L， et al. A novel hybrid decomposition-ensemble model based on VMD and HGWO for container throughput forecasting［J］. Applied Mathematical Modelling， 2018， 57： 163-178. 10.1016/j.apm.2018.01.014
24	DU P， WANG J Z， YANG W D， et al. Container throughput forecasting using a novel hybrid learning method with error correction strategy［J］. Knowledge-Based Systems， 2019， 182： No.104853. 10.1016/j.knosys.2019.07.024
25	LI H T， BAI J C， LI Y W. A novel secondary decomposition learning paradigm with kernel extreme learning machine for multi-step forecasting of container throughput［J］. Physica A： Statistical Mechanics and its Applications， 2019， 534： No.122025. 10.1016/j.physa.2019.122025
26	DING M， ZHOU H， XIE H， et al. A time series model based on hybrid-kernel least-squares support vector machine for short-term wind power forecasting［J］. ISA Transactions， 2021， 108： 58-68. 10.1016/j.isatra.2020.09.002
27	JIN F， LI Y W， SUN S L， et al. Forecasting air passenger demand with a new hybrid ensemble approach［J］. Journal of Air Transport Management， 2020， 83： No.101744. 10.1016/j.jairtraman.2019.101744
28	RICHMAN J S， MOORMAN J R. Physiological time-series analysis using approximate entropy and sample entropy［J］. American Journal of Physiology — Heart and Circulatory Physiology， 2000， 278（6）： H2039-H2049. 10.1152/ajpheart.2000.278.6.h2039
29	李文武，石强，李丹，等.基于VMD和PSO-SVR的短期电力负荷多阶段优化预测［J/OL］.中国电力. （2020-12-21）［2021-03-21］. . 10.1061/9780784482858.019
	LI W W， SHI Q， LI D， et al. Multi-stage optimization forecast of short-term power load based on VMD and PSO-SVR［J/OL］. Electric Power. （2020-12-21）［2021-03-21］. . 10.1061/9780784482858.019

[1]	徐萌, 王亚锟. 基于双向长短期记忆网络的DA40飞机碳刹车片剩余寿命预测[J]. 计算机应用, 2021, 41(5): 1527-1532.
[2]	朱林奇, 张冲, 周雪晴, 魏旸, 黄雨阳, 高齐明. 融合深度置信网络与与核极限学习机算法的核磁共振测井储层渗透率预测方法[J]. 计算机应用, 2017, 37(10): 3034-3038.

基于优化变分模态分解和核极限学习机的集装箱吞吐量预测

Container throughput prediction based on optimal variational mode decomposition and kernel extreme learning machine

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 29

相关文章 2

编辑推荐

Metrics